Exploring the link between National Institutes of Health (NIH) Toolbox Cognition Battery performance and everyday functioning in mild cognitive impairment.

Objective: Neuropsychologists routinely use scores on traditional paper-and-pencil tests to assess capacity for independent functioning, with these assessments accounting for a moderate amount (20%-37%) of the variance in instrumental activities of daily living (IADL) performance. The field is shifting toward incorporation of computerized neuropsychological assessments such as the National Institutes of Health Toolbox Cognition Battery (NIHTB-CB). There have been no studies examining how the NIHTB-CB relates to IADL performance or whether it better predicts IADL performance compared to traditional methods.

Method: Data from 74 participants in the Lithium as a Treatment to Prevent Impairment of Cognition in Elders were analyzed. Participants completed a neuropsychological assessment battery, the NIHTB-CB, and measures of IADL capacity. Linear regression was used to determine whether NIHTB-CB performance predicted IADL performance. Hierarchical multiple regression was used to determine whether NIHTB-CB added incremental validity to the prediction of IADL performance.

Results: Select NIHTB-CB measures predicted IADL functioning as measured by a performance-based assessment of everyday functioning, but not self- or informant-reported everyday functioning. Of these subtests, two were found to add incremental validity to the prediction of IADL performance above and beyond their matched traditional measures, with effect sizes ranging from mild to moderate (Δ² = .045-.077).

Conclusions: We found performance on individual NIHTB-CB subtests predicted approximately the same amount of variance in IADL functioning as traditional tests, with only subtests assessing processing speed and executive functioning adding a mild to moderate amount of additional variance above matched measures. It remains important for clinicians to consider additional determinants of functioning when using computerized assessments. (PsycInfo Database Record (c) 2025 APA, all rights reserved).